General > General Technical Chat
"Veritasium" (YT) - "The Big Misconception About Electricity" ?
rfeecs:
Yet another simulation video:
This one using a homebrew FDTD simulation with MATLAB. :clap:
daqq:
If anyone wants to play around with FDTD simulators without going to the bank, see openEMS. Effectively non-existent user interface, but works, I simulated some really nasty stuff with it.
https://openems.de/start/
adx:
--- Quote from: rfeecs on January 11, 2022, 07:30:52 pm ---This one using a homebrew FDTD simulation with MATLAB. :clap:
--- End quote ---
I didn't know FDTD was in Wikipedia, probably searched and just missed it. I'm fond of lashing up simulators using Yee lattice-like constructs (naively - just from my head, and see how it works) so was thinking about it. I wanted to probe and plot the same things, Silicon Soup has done a very good job, and properly-er than I ever would.
I can now confidently say I was wrong to think only 1/4 of the voltage across each transmission line is lost to radiation or whatever I said (based on a 'heuristic' guess that the un-driven wire in each leg is gently loaded while the driven ends are driven hard, such that what would be 1/2 loss guestimatically reduces to 1/4). Silicon Soup simulates the impedance match case, and shows the 1/c current is definitely way smaller than expected from transmission line theory, confirming COMSOL, AlphaPhoenix's and Ben Watson's videos.
And something else, I'll have to watch the video properly.
adx:
--- Quote from: adx on January 11, 2022, 04:02:52 am ---
--- Quote from: SandyCox on January 10, 2022, 01:59:52 pm ---According to transmission line theory, the bulb current is constant between t=0ns and t=100ns. So, the COMSOL simulation is delayed by 3.3ns compared to transmission line theory. I attach the full explanation.
Edit: I updated the note which now includes a comparison over the first 500ns.
--- End quote ---
Ah your note takes me back - is the sort of thing I have not done since uni (decades). Nothing personal against Laplace, but I found the transforms and representation extremely cumbersome - unbelievably so, I went to university believing I knew a fair bit about electronics as a hobby and knowing things like Ohm's law, so this was surprising at the time. It would have been extremely useful in the 1700s - 1800s. Z transform made a little more practical sense to me, closer to actual numbers. Oh - looking at Wikipedia, it was originally (due to Laplace) the discrete Z-transform. All this stuff I didn't know.
It might still be good to see it shifted 3.33ns (just an idea, not really necessary).
--- End quote ---
Reading this back, rather than silly it sounds like I'm being an arse - the latter not my intention. It's just that Laplace transforms were not for me, but I liked and understood it for possibly the first time in my life.
In fact I've learned more about the fundamental nature of electricity from this thread (and a few YT comments before I got here) than years of schooling and university. I don't think it's necessarily because of bad education (at either end), but that this puzzle has been a genuine challenge which lit up the world, inspiring people to tackle it in a variety of ways with some novel thinking and actions and questioning - in that sense this so-called thought experiment might just be more realistic than so much of going through the motions that technology has become.
To that end, there's still a thing that I don't understand. Without intending to cause a flood of exertion, new post:
adx:
Is surface charge real?
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version